1. Field of the Invention
This invention relates to systems and methods for adapting the output speed of a device to the speed of the output medium. More particularly, this invention relates to systems and methods for adapting the output rate of an Ethernet network device to the speed of the line to which it is connected.
2. Description of the Related Art
Over the last several years, the proliferation of the Internet has had a significant impact on many industries, especially the computer industry. The Internet has grown into an enormous network to which virtually any large or small computer network may be connected. In fact, it is now commonplace for the vast majority of people in free, industrialized nations to have access to the Internet, either through their business or work and/or through personal accounts. This connectivity has allowed many businesses, universities, governments, etc. to expand to provide their services via the Internet.
Most people or businesses receive their Internet access via an Internet Service Providers (ISPs). ISPs provide access to the Internet for their customers usually through membership subscriptions. ISPs make at least a portion of their income on service fees such as subscription fees, on-demand provisioning of services, etc. ISPs often regulate the amount of Internet bandwidth (i.e., data speed) that a customer is entitled.
Today's networks primarily support data services. However, new applications, such as network games, voice telephony and live video will elevate the data network infrastructure as an integral component of the next generation network. A typical home in the near future could have two phones, one or two personal computers and/or Personal Digital Assistants (PDA's) and general television sets. Applications will strain network bandwidth availability unless end-to-end quality of service (QOS) is available to guarantee real time voice and video quality as well as data integrity.
QOS can be affected by flow control issues and line mismatch that can cause delays in packet delivery and unintended congestion problems. For example, referring to FIG. 1, two devices are shown connected by a connection medium such as fiber, coaxial cable, twisted pair, etc. The network devices may be, for example, network switches, routers, hubs, etc. Device 1 includes a buffer 106a, a media access controller (MAC) 104a, and a physical interconnect (PHY) 102a. When the buffer of device 1 is full, the ingress rate control subsystem (e.g., ARL) of the device 1 MAC will send a PAUSE frame to prevent device 2 from sending additional frames. This protocol provides what is called back pressure to the traffic flow. In the meantime, device 2 may still be sending frames before the device 2 MAC receives the PAUSE frame. Device 1, therefore, will need to reserve enough buffer space to account for the line delay, which is based on the type of medium and the length of the line. Furthermore, each PHY and each device MII layer (not shown) may also add delay to the system. For example, a Reed-Solomon (R/S) interleaver may be included in the MII layer to interleave data packets and perform error correction. The R/S interleaver therefore adds delay to data packets traveling between the MAC and the PHY. This delay may be variable based on the size and type of data being interleaved.
Rate mismatch may occur especially when a VDSL PHY is connected to an Ethernet port. As an attempt to solve rate adaptation issues, some network device implementations use a buffer in the PHY to account for the rate mismatch. For example, referring to FIG. 2, shown is a network device which includes a buffer 106, a MAC 104 and a PHY 102. In between the PHY 102 and the MAC 104 in the physical layer are an input buffer 108a and an output buffer 108b. The input buffer 108a can account for the rate mismatch in outputting data from the PHY 102 to the medium, and the output buffer 108b can account for the rate mismatch between the PHY 102 and the MAC 104. In this proposed scenario, when the input buffer 108a is becoming full, a carrier sense (CRS) signal may be sent to the MAC 104 to defer transmission to the input buffer 108a. When a carrier sense (CRS) signal is asserted, the output buffer 108b can still continue to empty the buffer. Of course, in this scenario, the MII must be working in half duplex mode. However, problems may occur when there is an actual collision. When there is an actual collision, the MAC will back off sending packets, and part of the packet that is currently being sent becomes a fragment. In the meantime, the packet transmitted by the buffer is anticipated to be a fragment and gets dumped by the MAC. When the MAC dumps this packet, the packet will be lost since the buffer has not backed off and resent the packet. PHY to PHY backpressure is also needed (as well as buffer back-off), if the switch is rate controlling the port. Since in this mode the PAUSE frame is not available, the only mode of operation to avoid a subsequent collision is through jamming of the line.
Jamming is intended to cause the colliding stations to reschedule their next transmission attempts at different intervals in order to avoid subsequent collisions. Note that the jamming will have to pass through the buffer in the PHY, and there will be added latency caused by the buffer. To prevent large numbers of packet losses, a larger buffer must be used. However, a large buffer creates even larger latency. Additional control between the PHY and the MAC might be required.
In view of the problems described above, a solution is desired for adapting to the line rate and accounting for line delay and delays in the MII layer which does not include PHY buffers. Accordingly, there is a need for new systems and methods for rate adaptation in a data network.